Hsp90 inhibition protects against biomechanically

ARTHRITIS & RHEUMATISM
Vol. 65, No. 8, August 2013, pp 2102–2112
DOI 10.1002/art.38000
© 2013, American College of Rheumatology
Hsp90 Inhibition Protects Against
Biomechanically Induced Osteoarthritis in Rats
Michiel Siebelt,1 Holger Jahr,1 Harald C. Groen,1 Marjan Sandker,1 Jan H. Waarsing,1
Nicole Kops,1 Cristina Müller,2 Willem van Eden,3 Marion de Jong,1 and Harrie Weinans4
tor treatment increased cartilage sulfated glycosaminoglycan levels to concentrations even beyond baseline and
protected against cartilage degradation, stimulated subchondral bone thickness, and suppressed macrophage
activation.
Conclusion. Our findings indicate that Hsp90
plays a pivotal role in biomechanically induced chondrocyte stress responses. Intervention strategies that
inhibit Hsp90 can potentially protect or improve cartilage health and might prevent OA development.
Objective. Although articular cartilage has
evolved to facilitate joint mobilization, severe loading
can induce chondrocyte apoptosis, which is related to
the progression of osteoarthritis (OA). To avoid apoptosis, chondrocytes synthesize heat-shock proteins
(HSPs). This study was undertaken to examine the roles
of Hsp70 and Hsp90 in biomechanically induced OA,
and the possibility of using Hsp90 inhibition as an
intervention strategy for OA management.
Methods. OA was biomechanically induced in rats
by means of strenuous running. Disease progression
was compared between running rats treated with Hsp90
inhibitor and untreated running controls. Hsp70 and
Hsp90 protein levels in articular cartilage were determined by Western blotting. OA progression was monitored
using contrast-enhanced micro–computed tomography
to measure cartilage degradation and subchondral bone
changes and single-photon–emission computed tomography to examine synovial macrophage activation and
histologic features.
Results. Chronic cartilage loading led to early OA
development, characterized by degeneration of cartilage
extracellular matrix. In vivo Hsp90 inhibition resulted
in increased Hsp70 synthesis, which suggests that
Hsp90 activity limits Hsp70 production. Hsp90 inhibi-
Osteoarthritis (OA) is a progressive disease affecting synovium, ligaments, and subchondral bone, and
can lead to articular cartilage degradation (1). Approximately 30% of persons ages ⱖ65 years are affected by
this severely disabling disease in the hip or knee joint
(2). Besides costly and invasive joint replacement surgery, treatment options are limited. Therefore, more
detailed knowledge of the underlying pathogenesis of
OA is essential for the development of diseasemodifying drugs.
Chondrocytes are the single cell type responsible
for maintaining the extracellular matrix (ECM) of articular cartilage and repair of any inflicted damage. However, being exposed daily to high-peak forces during
physical activity, this cell type is sensitive to mechanical
stimuli (3). Acute or chronic high-intensity loads can
cause cartilage damage (4), and chondrocytes in damaged or eroded cartilage show morphologic features of
apoptosis, suggesting that chondrocytes from OA patients die by active (programmed) cell death (5). Apoptosis can be prevented by expression of heat-shock
proteins (HSPs). HSPs are molecular chaperones that
assist in protein folding to sustain cellular homeostasis
under stressed conditions (6).
Hsp70 and Hsp90 are 2 of the major classes of
HSPs involved in the regulation of cell stress (7). From
in vitro experiments it is known that Hsp70 inhibits nitric
oxide–induced apoptosis in chondrocytes through a re-
Supported by the Dutch Arthritis Association, the Netherlands Ministry of Economic Affairs (BMM/TerM P2.02 Program), and
the Netherlands Ministry of Education, Culture, and Science.
1
Michiel Siebelt, MD, Holger Jahr, PhD, Harald C. Groen,
PhD, Marjan Sandker, MD, Jan H. Waarsing, PhD, Nicole Kops, BSc,
Marion de Jong, PhD: Erasmus Medical Center, Rotterdam, The
Netherlands; 2Cristina Müller, PhD: Center for Radiopharmaceutical
Sciences ETH/PSI/USZ and Paul Scherrer Institute, Villigen, Switzerland; 3Willem van Eden, PhD: Utrecht University, Utrecht, The
Netherlands; 4Harrie Weinans, PhD: Erasmus Medical Center, Rotterdam, The Netherlands, Delft University of Technology, Delft, The
Netherlands, and UMC Utrecht, Utrecht, The Netherlands.
Address correspondence to Michiel Siebelt, MD, Department
of Orthopedics, Erasmus Medical Center, PO Box 2040, 3000 CA
Rotterdam, The Netherlands. E-mail: m.siebelt@erasmusmc.nl.
Submitted for publication November 12, 2012; accepted in
revised form April 26, 2013.
2102
Hsp90 IN OSTEOARTHRITIS
2103
Figure 1. Representation of hypothetical roles of Hsp70 and Hsp90 in osteoarthritis (OA). A, Healthy cartilage. Normal chondrocytes reside in a
sulfated glycosaminoglycan (sGAG)–rich extracellular matrix (ECM). Subchondral bone is intact, and supportive, inactive macrophages are present
within the synovium. B, Stressed cartilage during strenuous running. A small amount of sGAG is lost from the ECM. Due to loss of cartilage
hydrostatic pressure, chondrocytes become biomechanically stressed. In order to cope, chondrocytes up-regulate Hsp70. C, Persisting biomechanical
stress. Stressed chondrocytes have diminished biosynthetic capacity and produce less sGAG, and Hsp90 is up-regulated. Activated osteoclasts
penetrate the subchondral plate, impairing its supportive function. Macrophages produce proinflammatory cytokines and growth factors. D, OA.
Eventually, a vicious circle of events, as described in C, culminates in chondrocytes ultimately dying from apoptosis. The integrity of the ECM is
compromised and its biomechanical properties deteriorate. Osteoclasts tunnel their way through the subchondral bone, making space for osteoblast
and vascular infiltration, which will eventually lead to the development of a sclerotic bone phenotype. Continuous macrophage cytokine and growth
factor production thickens the synovium and results in fibrosis, reducing the patient’s range of motion and causing pain. Color figure can be viewed
in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.38000/abstract.
duction in caspase 3 activity (8). Galois et al reported
similar findings in an in vivo study of rats and showed
that mild running (7.5 km in 28 days) and moderate
running (15 km in 28 days) stimulated Hsp70 production
(9). These schedules protected chondrocytes against
caspase 3–induced apoptosis and reduced OA progression. Other studies further support the finding that
Hsp70 has the potential to prevent cartilage damage in
arthritic joints (10,11). Thus, Hsp70 is thought to play a
protective role in early stages of chondrocyte adaptation
to biomechanical joint constraints, which otherwise
would lead to OA (12). However, intense running (30
km in 28 days) reduced Hsp70 expression back to
nonrunning control levels (9) and, therefore, supraphysiologic loading seems to exceed the intrinsic Hsp70mediated capacity for cellular damage control, ultimately still resulting in apoptosis. This corroborates
findings that intense or strenuous running protocols are
known to induce OA (13,14). It is hypothesized that with
persistent stress on cartilage, either increased levels of
Hsp70 are insufficient to protect chondrocytes (15), or
the coexpression of other HSPs counteract the effect of
Hsp70 (9).
The second part of this hypothesis could be
explained through Hsp90 function. Whereas Hsp70 inhibits NF-␬B formation, Hsp90 has an antagonistic
function and activates the NF-␬B pathway (6). Although
NF-␬B plays an essential role in normal physiology,
inappropriate regulation of NF-␬B is related to the
pathogenesis of both OA and rheumatoid arthritis (RA)
(16). In an in vivo model of RA, Hsp90 inhibition
reduced the inflammatory response, prevented cartilage
damage, and limited bone resorption (17). In addition,
Hsp90 may restrict Hsp70 regulation (18) and, therefore, Hsp90 up-regulation in chondrocytes may counteract the beneficial Hsp70-mediated responses and facilitate OA progression. Hsp90 inhibition might positively
affect the level of Hsp70 up-regulation necessary and,
through this mechanism, promote cartilage health.
In summary, Hsp70- and Hsp90-related chondrocyte stress responses might play an important role in
biomechanically induced cartilage stress that develops
into OA (Figure 1). Increased physical activity in a
model of OA induced by strenuous running exposes
chondrocytes in healthy cartilage to chronic impact from
joint loading (Figure 1A). To cope with this increased
load, chondrocytes may up-regulate Hsp70 to promote
cell metabolism to sustain the ECM with sufficient
sulfated glycosaminoglycan (sGAG) (Figure 1B). However, persisting cartilage stress through strenuous running could stimulate Hsp90 production, which counteracts the positive effects of Hsp70 and thus promotes OA
development (loss of sGAG, degradation of ECM, loss
of subchondral bone, and activation of macrophages)
(Figure 1C). When cartilage ECM is damaged, OA
2104
SIEBELT ET AL
progression will further develop and, in time, progress to
severe joint degeneration (Figure 1D).
The present study investigates the role of Hsp70
and Hsp90 in a rat model of OA induced by strenuous
running. We hypothesized that running-induced loading
of articular cartilage up-regulates Hsp90 expression in
chondrocytes and limits Hsp70 expression, which results
in a loss of the Hsp70-mediated protective effect against
OA progression in the knee joint. Therefore, inhibition
of Hsp90 might reverse the degenerative effects on
cartilage after strenuous running in rats.
MATERIALS AND METHODS
Study design. Thirty-eight 16-week-old male Wistar
rats (Charles River Netherlands) were housed at 21°C on a
12-hour light/dark cycle during the experimental period. Male
rats were used in this study since estrogen is known to
influence both HSP expression (19) and OA development (20).
Animals received standard food pellets and water ad libitum.
The following 5 groups were formed: a baseline group
(n ⫽ 6), 2 control OA groups, 1 that was followed up for 6
weeks (n ⫽ 6) and 1 that was followed up for 12 weeks (n ⫽ 6),
and 2 treatment groups (both treated with the Hsp90 inhibitor
BIIB021), 1 that was followed up for 6 weeks (n ⫽ 10) and 1
that was followed up for 12 weeks (n ⫽ 10).
All rats (except the baseline group) were run on a
motorized rodent treadmill (LE-8700; Panlab Harvard Apparatus). In Wistar rats, strenuous running induces the development of OA through chronic mechanical loading of articular
cartilage (14). After an initial training week to acclimate rats
to the treadmill exercise, rats were run 1 hour per day, 5 days
per week (not on weekends). During the first 10 minutes, the
rats ran 20 cm/second (0.72 km/hour) to warm up. This was
followed by 50 minutes of running at 33.3 cm/second (1.20
km/hour). The pace and duration of this protocol are equal to
⬃50% of a total exhaustion protocol (21). At the end of the
6-week protocol, the rats had completed 30 km of strenuous
running, which leads to the development of moderate OA in
their knee joints that does not heal spontaneously (13).
We measured changes in Hsp70 and Hsp90 levels in rat
articular cartilage to characterize chondrocyte stress responses
and relate this to OA development. OA is characterized by
sGAG loss from cartilage ECM, cartilage ECM loss, bone
resorption, and macrophage activation. For each group, at
the end of the experiment, an analysis sequence of micro–
computed tomography (micro-CT) and micro–single-photon–
emission computed tomography used in combination with
micro-CT as an anatomic reference (SPECT/CT) was performed to measure changes in these different aspects of OA.
This sequence consisted of in vivo 111In–diethylenetriaminepentaacetic acid (DTPA)–folate SPECT/CT to measure macrophage activation, ex vivo equilibrium partitioning of an ionic
contrast agent (EPIC) micro-CT to measure bone and cartilaginous tissue changes, Western blot analysis of articular
cartilage Hsp90 and Hsp70, and histologic analysis to examine
cartilaginous tissue quality.
Figure 2 presents a detailed planning scheme of all
groups and conducted tests. All animal protocols were ap-
Figure 2. Experimental overview indicating analytical time points and
methods for each group of animals. Male Wistar rats (age 16 weeks;
n ⫽ 38) were divided into 3 groups: a baseline group (n ⫽ 6), 2
untreated groups with osteoarthritis (OA; n ⫽ 12), and 2 groups with
OA treated with the oral Hsp90 inhibitor BIIB021 (n ⫽ 20). All rats,
with the exception of those in the baseline group, were subjected to a
6-week strenuous running protocol to induce OA (13). One group of
untreated rats with OA (n ⫽ 6) and one group of rats with OA treated
with Hsp90 inhibitor (n ⫽ 10) underwent a subsequent 6 weeks of rest
before analysis. At the end of the experiment, a full analysis sequence
was performed, consisting of in vivo measurement of activated macrophages using 111In–diethylenetriaminepentaacetic acid–folate singlephoton–emission computed tomography, ex vivo bone and cartilage
analysis using equilibrium partitioning of an ionic contrast agent
micro–computed tomography, histologic analysis, and quantification
of Hsp70 and Hsp90 levels by Western blotting.
proved by the Animal Ethics Committee of Erasmus Medical
Center.
Hsp90 inhibition via orally administered BIIB021. All
animals in both treatment groups were treated orally with the
fully synthetic Hsp90 inhibitor BIIB021 (Selleck Chemicals),
which competitively binds in the ATP-binding pocket of
Hsp90, similar to other geldanamycin-derived Hsp90 inhibitors
(22). Previous work has shown that Hsp90 inhibitors do not
interfere with Hsp70 functioning (23). BIIB021 was dissolved
in DMSO and further diluted in saline to a 0.02% DMSO
solution with a BIIB021 concentration of 14 mg/ml. Throughout the entire study, each animal received 0.5 ml of this
solution with each dose of BIIB021 via oral probing. High
doses of other types of Hsp90 inhibitors (such as geldanamycin
derivatives) result in drug-related gastrointestinal, bone marrow, or hepatic toxicities (24). Therefore, each animal received
7 mg of BIIB021 3 times per week, every other day except on
weekends, so that cells in the gastrointestinal tract and bone
marrow could recover from Hsp90 inhibition.
Detection of activated macrophages by SPECT/CT
using 111In-DTPA-folate. Activated macrophages express folate receptor ␤, allowing the monitoring of macrophages in
vivo using folate radioconjugates (25). This technique was
recently introduced for OA research in a rat model (26).
Briefly, DTPA-folate (conjugate kindly provided by Professor
R. Schibli, Center for Radiopharmaceutical Sciences ETHPSI-USZ, Zurich, Switzerland) was incubated with 111InCl3
(Mallinckrodt-Tyco) in phosphate buffered saline (PBS; pH
6.5) for 1 hour at room temperature. Quality control performed using high-performance liquid chromatography revealed a radiochemical yield of ⬃92% at a specific activity of
Hsp90 IN OSTEOARTHRITIS
⬎16 MBq/␮g. After the addition of a solution of DTPA for
complexation of traces of free 111In(III), the solution was
further diluted in PBS and administered via the tail vein 20
hours prior to scanning.
SPECT/CT scans were performed with a 4-head multiplex multipinhole small animal SPECT/CT camera (NanoSPECT/CT; Bioscan). Each detector head was fitted with a
tungsten-based collimator of nine 2.5-mm diameter pinholes,
the field of view was 24 mm in width, and energy peaks were set
at 170 keV and 240 keV (⫾10%). All rat knee joints were
scanned using both helical micro-CT (acquisition time 5 minutes) and SPECT/CT (acquisition time 30 minutes).
After scanning, all data sets were reconstructed at an
isotropic CT voxel size of 0.2 mm3 and an isotropic SPECT/CT
voxel size of 0.6 mm3 using HiSPECT software (Scivis). All
scans were analyzed using InVivoScope processing software
(Bioscan). A cylindrical region of interest (ROI) was manually
determined for quantification of the radioactivity around the
knee joint; all data are presented as activity measured per
mm3.
Measurements of rat bone, cartilage, and growth plate
using EPIC micro-CT. OA is characterized by loss of sGAG
from the cartilage ECM, followed by cartilage degradation.
Cartilage and growth plate x-ray attenuation from contrast
(ioxaglate)–enhanced micro-CT scans (EPIC micro-CT) is
inversely related to the sGAG content of cartilage (27) and
indicative of tissue quality (13). With EPIC micro-CT, it is
possible to accurately quantify morphometric parameters of
cartilaginous tissue, as well as the subchondral bone (28).
Animals were killed immediately after the SPECT/CT
scan. Both knee joints were harvested and randomly assigned
to EPIC micro-CT or protein analysis by Western blotting. Soft
tissue was carefully removed to a maximal extent from all knee
joints selected for EPIC micro-CT, without harming cartilage
integrity. Next, all specimens were incubated in a 40% solution
of ioxaglate for 24 hours at room temperature (29). EPIC
micro-CT was performed on a SkyScan 1076 in vivo micro-CT
scanner, using the following scan settings: isotropic voxel size
35 ␮m, voltage 55 kV, current 181 mA, field of view 68 mm,
with a 0.5-mm aluminum filter, over 198° with a 0.4° rotation
step. All scans were performed using the same settings, and all
data were reconstructed identically. Using CT analysis software (SkyScan), these data sets were segmented using a fixed
attenuation threshold between air (at 25) and subchondral
bone (at 100) (13). In all segmented micro-CT data sets, ROIs
were drawn around the cartilage of the medial and lateral
plateau of the tibia to calculate x-ray attenuation (arbitrary
gray values), which is inversely related to sGAG content
(cartilage quality), and cartilage thickness (in micrometers).
Tibial plateau cartilage was analyzed since it is predominantly
affected during OA induced by strenuous exercise (30).
Bone was accurately segmented from all EPIC
micro-CT data sets using a local threshold algorithm (31).
Cortical and trabecular bone were automatically separated
using in-house software (32) (details regarding both 3-D
calculator software and separation software are available from
the corresponding author upon request). Using the CT analysis
software (SkyScan), the tibial epiphysis was selected in the
segmented CT scans and analyzed for changes in cortical and
trabecular bone. Both the medial and lateral thicknesses of the
subchondral plate were measured.
2105
Histopathologic examination of the rat knee joint.
After EPIC micro-CT, the separated parts of the rat knee
joints were fixed in paraformaldehyde, decalcified with formic
acid, and embedded in paraffin. Sagittal sections were cut at
300 ␮m intervals and stained with Safranin O to detect the
amount and distribution of GAG. All sections were stained in
parallel, to minimize staining bias between different samples.
The mid-sections of both the medial and lateral tibial plateau
were digitized with the NanoZoomer Digital Pathology program (Hamamatsu Photonics). From these digital images, the
cartilage was isolated in silico, and the staining intensity was
quantified using graphics software (Adobe Photoshop). Cartilage thickness was measured at 7 different locations using
NanoZoomer, and the mean thickness was recorded as the
average histologic cartilage thickness (33). Additionally, medial and lateral tibial plateau sections were scored according to
the OA Research Society International (OARSI) histopathology initiative (34).
Protein extraction and Western blotting. From all rat
knee joints selected for protein analysis, the articular cartilage
(both medial and lateral) of the tibial plateau was harvested,
rinsed in physiologic saline, immediately snap-frozen in liquid
nitrogen, and stored at ⫺80°C until used. The samples were
pulverized for 2 minutes at 30 Hz in a TissueLyser II (Qiagen)
using chromium steel grinding balls and custom-made polytetrafluoroethylene vials. Samples were resuspended in lysis
buffer (35) and purified as previously described (36). Protein
quantification, loading of sodium dodecyl sulfate–polyacrylamide
gels, blotting, and signal quantification were performed as
previously described (37), and Hsp70 and Hsp90 antigens were
detected using the methods described by Xing et al (38) and
Hirano et al (39), respectively, using the recommended antibody dilutions. Visualization was performed on an Odyssey
infrared imaging system with IRDye 680RD and IRDye
800CW secondary antibodies (1:15,000; both from Li-Cor
Biosciences), respectively. Replicate data per animal were
averaged and normalized to ␣-tubulin (1:1,000; Cell Signaling
Technology) as a loading control. Signal intensities were quantified using ImageJ software (National Institutes of Health).
Geldanamycin-like inhibitors that compete for the
N-terminal nucleotide-binding pocket interact most potently
with the cytoplasmic isoforms Hsp90␣ and Hsp90␤ (40,41).
Both isoforms were measured by Western blotting. BIIB021
interferes with the ATPase activity of Hsp90. However, because this method is unable to distinguish between the activation stages of the HSPs, the data presented represent total
HSP content (either inactive or active).
Statistical analysis. Differences between mean values
for animals in the baseline group and both groups of untreated
controls with OA were tested using one-way analysis of variance (ANOVA) with Bonferroni correction (for Western
blotting, EPIC micro-CT, and quantitative histologic analysis
data) (SPSS). Differences in semiquantitative histology scores
between animals in the baseline group and animals in both
groups of untreated control rats with OA were analyzed using
a Kruskal-Wallis one-way ANOVA. To compare the means of
quantitative outcome measurements (Western blotting, EPIC
micro-CT, and quantitative histologic analysis) at 6 or at 12
weeks between untreated animals with OA and animals treated
with Hsp90 inhibitor, an unpaired t-test was used. A MannWhitney test was used to compare the means of semiquantitative histology scores between untreated controls with OA
2106
SIEBELT ET AL
and animals treated with Hsp90 inhibitor. The amount of
injected radioactivity per animal can influence DTPA-folate
SPECT/CT macrophage measurements. In order to correct for
this possible influence when comparing differences between
groups, the amount of injected activity was added as a covariable in a linear regression model using SPSS. P values less than
0.05 were considered significant.
RESULTS
Hsp70 and Hsp90 regulation in OA and Hsp90
inhibition. After a 6-week regimen of running, Hsp90
protein levels were increased 2.1-fold in untreated control rats with OA compared to untreated animals in the
baseline group that did not run (Figure 3A). After 6
weeks of subsequent rest, Hsp90 levels in untreated
control rats with OA were still higher than those in the
baseline group (⬃1.7 fold increase) (Figure 3A). However, neither of these effects were significant (P ⫽ 0.12).
Compared to baseline values, Hsp70 levels did not
change at 6 weeks or 12 weeks in untreated control rats
with OA (P ⫽ 0.55) (Figure 3B).
Chondrocytes from animals treated with Hsp90
inhibitor showed a different HSP response to biomechanical stress exposure via running. Compared to untreated control rats with OA, Hsp90 inhibitor–treated
Figure 3. Changes in Hsp90 and Hsp70 abundance during chronic
loading and after Hsp90 inhibition in articular chondrocytes from rat
knee cartilage. A and B, Levels of Hsp90 (A) and Hsp70 (B) in rats that
were not subjected to the running protocol (0 weeks), untreated
control rats with running-induced osteoarthritis (OA), and rats with
running-induced OA treated with Hsp90 inhibitor (Hsp90i). Bars show
the mean and 95% confidence interval. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01.
C, Representative Western blots of Hsp90 and Hsp70 levels. ␣-tubulin
was used as a loading control.
animals produced higher amounts of Hsp90 at the end
of the running protocol (⬃1.9 fold increase; P ⫽ 0.02),
which remained increased after the subsequent 6 weeks
of rest (⬃2.2 fold increase; P ⫽ 0.001) (Figure 3A). The
increased levels of Hsp90 are consistent with a functional inactivation of Hsp90 by BIIB021, as earlier
studies demonstrated a positive feedback regulation of
this HSP (22). Further evidence in favor of a functional
Hsp90 inactivation can be derived from the subsequent
Hsp70 induction (18,22,42), as was observed in the
articular cartilage of rats after running (P ⫽ 0.04)
(Figure 3B). At week 12, Hsp70 protein levels in Hsp90
inhibitor–treated rats were still increased ⬃2.2 fold
compared to nonrunning animals in the baseline group
(P ⫽ 0.04), but were not significantly different from the
levels in untreated controls with OA (P ⫽ 0.10) (Figure
3B). Representative Western blot images are shown in
Figure 3C.
Adaptation of cartilaginous tissue. OA is characterized by a loss of sGAG from the cartilage ECM,
followed by cartilage degradation. Strenuous running in
untreated control rats with OA did not induce clear
changes in the articular cartilage of the medial tibial
plateau (Figures 4A and B), while at the lateral side,
cartilage did show a reduction in ECM thickness (P ⫽
0.02) (Figure 4D). Hsp90 inhibitor–treated animals had
lower attenuation values for both medial plateau (P ⫽
0.04) (Figure 4A) and lateral plateau (P ⫽ 0.02) cartilage (Figure 4C). This indicates that Hsp90 inhibitor–
treated rats had higher levels of sGAG to sustain the
cartilage during running. Not only was the sGAG content in both the medial and lateral compartments higher
compared to that in the untreated control rats with OA,
Hsp90 inhibitor–treated animals even had higher
amounts of sGAG compared to healthy animals in the
baseline group. After 6 weeks of rest, medial cartilage
from the Hsp90 inhibitor–treated animals still had
higher sGAG levels (P ⫽ 0.04), while lateral cartilage
showed no difference in sGAG content (P ⫽ 0.88)
between the Hsp90 inhibitor–treated animals and control animals with OA. However, whereas lateral cartilage
from untreated control rats with OA was degraded,
Hsp90 inhibitor–treated animals showed no sign of
decreased cartilage thickness, and cartilage was thicker
compared to that in the control rats with OA (P ⫽ 0.009)
(Figure 4D).
To further validate these findings, we quantitatively evaluated the sGAG content of cartilage and
cartilage thickness in histologic sections. These measurements showed similar patterns between untreated control rats with OA and Hsp90 inhibitor–treated animals
and confirmed our EPIC micro-CT results (Figures
Hsp90 IN OSTEOARTHRITIS
2107
Figure 4. Differences in cartilaginous tissue between untreated control rats with osteoarthritis (OA) and rats with OA treated with Hsp90 inhibitor.
A–J, Examination of cartilage and the epiphyseal growth plate using equilibrium partitioning of an ionic contrast agent (EPIC) micro–computed
tomography (micro-CT) (A–E) and histologic analysis (F–J). Both techniques were used to measure the amount of sulfated glycosaminoglycans
(sGAG) and cartilage extracellular matrix (ECM) thickness. Attenuation values from EPIC micro-CT scans were inversely related to sGAG content.
Bars show the mean and 95% confidence interval. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.001. K, Representative images of Safranin O–stained
sections of rat medial and lateral tibial plateau cartilage and the medial and lateral growth plate at baseline, after 6 weeks of running, and after a
subsequent 6 weeks of rest. Cartilage and growth plate showed loss of sGAG at 6 weeks in untreated animals. In contrast, increased sGAG content
was observed in Hsp90 inhibitor–treated animals. At 12 weeks, cartilage damage was evident in untreated animals. Arrows indicate matrix damage.
Cartilage damage was less severe in Hsp90 inhibitor (Hsp90i)–treated animals. During the 6 weeks of rest, sGAG content was restored in the growth
plates of untreated animals. Insets show the complete histologic sections of the proximal tibia from which the detailed sections of cartilage and
growth plate were isolated. L and M, Cartilage damage, as measured by the OA Research Society International (OARSI) score, in the medial plateau
(L) and lateral plateau (M) in untreated and Hsp90 inhibitor–treated animals. Data are shown as box plots, where the boxes represent the 25th to
75th percentiles, the lines within the boxes represent the median, and the lines outside the boxes represent the maximum and minimum values.
Symbols inside the boxes represent the mean. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. Color figure can be viewed in the online issue, which is available at
http://onlinelibrary.wiley.com/doi/10.1002/art.38000/abstract.
2108
Figure 5. Subchondral bone thickness (sub. chond. plate th.), as
measured by micro–computed tomography (micro-CT), of the medial
plateau (A) and lateral plateau (B) in rats that were not subjected to
the running protocol (0 weeks), in untreated control rats with runninginduced osteoarthritis (OA), and in rats with running-induced OA
treated with Hsp90 inhibitor. Bars show the mean and 95% confidence
interval. ⴱ ⫽ P ⬍ 0.05.
4F–I). Additional semiquantitative scores according to
the OARSI histopathology initiative showed increased
medial (P ⫽ 0.002) and lateral (P ⫽ 0.012) cartilage
degeneration in untreated controls with OA. In contrast,
Hsp90 inhibitor–treated animals showed less cartilage
degeneration in both anatomic regions (Figures 4L–M).
Representative histology images of cartilage are shown
in Figure 4K.
The growth plate is a cartilaginous tissue in which
bone is formed via endochondral ossification. In rats, the
SIEBELT ET AL
growth plate never closes and remains highly chondral
throughout their lifespan. Increased activity through
treadmill running showed that the growth plate is sensitive to high impact from joint loading. EPIC micro-CT
and histology revealed that, after strenuous running, the
growth plate was severely depleted of sGAG (Figures 4E
and J). However, Hsp90 inhibition prevented this sGAG
loss. Representative histology images of the growth plate
are shown in Figure 4K.
Subchondral bone changes. OA is also characterized by periarticular bone changes, which we evaluated
with EPIC micro-CT scans. In untreated rats with OA,
the medial subchondral plate thickness slowly increased
during the study (⬃10% increase after 12 weeks, P ⫽
0.007). Hsp90 inhibitor–treated animals showed a faster
response. Their subchondral bone was already thicker
than that of untreated control rats with OA at 6 weeks
(P ⫽ 0.04) (Figure 5A). This increase did not progress
over time, and Hsp90 inhibitor–treated animals had a
similar subchondral bone thickness compared to untreated controls with OA at 12 weeks. At this time point,
the lateral subchondral bone of untreated controls with
OA showed a different response and was ⬃7% thinner
than that of Hsp90 inhibitor–treated animals (P ⫽ 0.04)
(Figure 5B).
Macrophage activation. OA-related macrophage
activation was measured in vivo with 111In-DTPA-folate
SPECT/CT. Animals in all groups received a mean ⫾ SD
of 82 ⫾ 5 MBq of 111In-DTPA-folate. Relative to
animals in the baseline group, there was only a slight
trend in macrophage activation in untreated control
Figure 6. Macrophage activation determined using single-photon–emission computed tomography (SPECT/CT) 20 hours after injection of
111
In–diethylenetriaminepentaacetic acid–folate (111In-DPTA-Fa) into the rat tail vein. A, Measured radioactivity in the knee joints of untreated rats
with osteoarthritis (OA) and rats with OA treated with Hsp90 inhibitor. Measurements were corrected for analyzed volume (mm3). High
radioactivity is related to greater macrophage activation. Bars show the mean and 95% confidence interval. ⴱⴱ ⫽ P ⬍ 0.01. B, Sagittal SPECT/CT
images of knee joints from representative animals in each experimental group. CT images are shown in black and white and were used for anatomic
reference; SPECT/CT images are shown in color.
Hsp90 IN OSTEOARTHRITIS
animals with OA after 6 weeks of running (P ⫽ 0.1). The
amount of activated macrophages present in Hsp90
inhibitor–treated animals after 6 weeks of strenuous
running was significantly lower than in untreated control
rats with OA (P ⫽ 0.008) (Figure 6). In Hsp90 inhibitor–
treated animals, macrophage activation levels showed a
clear increase from 6 to 12 weeks (P ⬍ 0.0001).
DISCUSSION
HSPs evolved to protect cells against physiologic
stress. Hsp90 directly promotes cell survival through
formation of active NF-␬B (6), indirectly promotes cell
survival via Hsp90–Akt complexes that inhibit JNKmediated cell death through phosphorylation and consequent inactivation of apoptosis signal–regulating kinase 1 (43), modulates the intrinsic pathway of apoptosis
by inhibiting oligomerization of Apaf-1 (44), and influences glucocorticoid receptor, shaping cellular responses
to glucocorticoids (45). However, this system may become overwhelmed under high-end stresses, and then
tissue homeostasis is lost (46). In the present study, we
used strenuous running as an established method to
induce a mild OA phenotype in rats (13) (Figure 4).
Running caused a trend toward Hsp90 accumulation in
cartilage that did not decline and remained elevated
during 6 weeks of rest, while Hsp70 levels remained
unaltered (Figure 3). Interestingly, it is known that
during a subsequent period of rest after strenuous
exercise, OA continues to progress, lacking spontaneous
cartilage repair (13).
In contrast to previous reports of increased
Hsp70 levels after 28 days of exercise (9), we measured
Hsp70 and Hsp90 only after 42 and 84 days. Due to
analysis at these limited time points, we are unable to
conclude what specific mechanism leads to the changes
in Hsp70 and Hsp90 levels. To identify strenuous running as an HSP inducer, more studies are needed that
focus on earlier time points, and additional samples
should be analyzed immediately after running and after
several hours of rest. In this model, we used BIIB021
treatment to evaluate potential OA-modifying aspects
of pharmacologic Hsp90 inhibition. Systemically introduced Hsp90 inhibition may be toxic (24). However, our
treatment regimen did not result in weight or hair loss in
the treated animals, which might have suggested Hsp90
inhibition–related toxicity.
Hsp90 inhibition resulted in elevated Hsp90 protein levels as well as higher Hsp70 levels. Previous work
with BIIB021 and other Hsp90 inhibitors clearly demonstrated a concentration-dependent effect of Hsp90
inhibition on the induction of both Hsp90 and Hsp70.
2109
Although elevation of Hsp90 levels seems to be a
counterintuitive result of Hsp90 inhibition, it is a wellknown response to different Hsp90 inhibitors, such as
geldanamycin, 17-allylamino-17-demethoxygeldanamycin,
and BIIB021. Through binding in the ATP-binding
pocket of Hsp90, they prevent activation of Hsp90,
which results in reduced Hsp90 activity with degradation
of downstream client proteins, such as human epidermal
growth factor receptor 2, AKT, and Raf-1 (22,47).
Functional inactivation of Hsp90 by BIIB021 also
induces Hsp70, which can be explained by the role of
heat-shock factor 1 (HSF-1). HSF-1 plays an established
role in the regulation of Hsp70 levels. It is kept in a
latent state by a stress–protein complex and is activated
upon proteotoxic insults (overloading) in order to activate Hsp70 gene expression (48). Hsp90 is a major
repressor of HSF-1 gene expression and retains HSF-1
in an inactive nontrimeric state (49). When mechanical
loading increases, Hsp90 production is increased, which
reduces HSF-1 activity, and Hsp70 up-regulation is
prevented. This inverse regulation of Hsp70 and Hsp90
through HSF-1 may explain why Hsp70 is known to be
less responsive to increased loads (50) and why the
protective effect of Hsp70 for maintaining ECM homeostasis and cartilage protection is compromised in our
running-induced model of OA. In the present study,
Hsp90 inhibitor–treated animals showed increased
Hsp70 protein levels. As one would expect, Hsp90
inhibition strongly induces HSF-1 in human chondrocytes (51). This explains why diminished Hsp90 activity
shifts the balance in favor of Hsp70 synthesis (18,22,42)
and stimulates the Hsp70 protective effect on cartilage.
It is due to this feedback loop that Hsp70 is used as a
standard pharmacodynamic biomarker for the analysis
of Hsp90 inhibitors functioning in both preclinical and
clinical studies (52).
However, Hsp90 inhibition can also directly influence cellular processes that reduce OA progression.
Hsp90 inhibition of in vitro–cultured human articular
chondrocytes selectively inhibited interleukin-1␤–induced
ERK activation and resulted in reduced matrix metalloproteinase 13 (MMP-13) production (53). Since excessive MMP-13 activity results in articular cartilage degradation, reduced MMP-13 production might prevent OA
(54). The findings of the present study do not allow us
to establish whether Hsp90 inhibition reduced OA progression directly or indirectly via increased levels of
Hsp70. However, the final outcome of Hsp90 inhibition
was marked changes in cartilage degradation, subchondral bone remodeling, and synovial macrophage activation.
In the present study, articular cartilage from
2110
Hsp90 inhibitor–treated animals had higher amounts of
sGAG (Figure 4), which probably gives cartilage the
necessary hydrostatic stiffness to absorb impact during
running and to protect chondrocytes against increased
mechanical stress. Hsp90 inhibitor treatment also resulted in a more prompt increase in medial subchondral
bone thickness compared to untreated control rats with
OA, and Hsp90 inhibition prevented subchondral bone
plate thinning in the lateral compartment (Figure 5).
Strenuous running did not induce macrophage activation in untreated animals. Nevertheless, animals treated
with Hsp90 inhibitor did show reduced levels of macrophage activation after 6 weeks of running (Figure 6).
However, despite continuous Hsp90 inhibitor treatment,
macrophage activation increased again during the subsequent 6 weeks of rest (Figure 6) and may suggest that
Hsp90 inhibitor did not directly modulate macrophage
responses.
A direct translation of our results into a clinical
treatment for OA patients may not be possible. BIIB021
and other geldanamycin-derived Hsp90 inhibitors are
currently used in cancer trials. Treatment of lifethreatening diseases might justify higher risks of Hsp90associated dose-limiting toxicities (24), but can never
be accepted in OA patients. Therefore, more detailed
knowledge of the downstream targets that are modulated by Hsp90 inhibition is needed. Another way to
reduce systemic side effects is to investigate whether
local treatment via intraarticular injections with Hsp90
inhibitor is feasible and beneficial.
Hyaline articular cartilage evolved to absorb
forces that develop during joint mobilization. From this
perspective, a balance can be expected between biomechanical loads and chondrocyte functioning, and an
imbalance is likely to result in cartilage failure and OA
development. When stress on cartilage is increased,
either via increased loading or due to changed joint
biomechanics (e.g., as a result of ligament tears) (9),
chondrocytes up-regulate HSPs, which suggests a pivotal
role in OA onset. However, few studies report HSP
production by chondrocytes as a possible regulator of
cartilage homeostasis under stressed conditions. More
research on this topic will lead to a more accurate
explanatory model for pathologic joint loading–induced
OA. Ambivalent effects of training on cartilage are
well known in clinical patient care (55). A combined
approach of regulated physical exercise and therapeutic
intervention of HSP production might reduce stress
exposure of chondrocytes in OA patients.
The results of our in vivo study strongly suggest
that chondrocyte stress-induced Hsp90 synthesis plays
an important role in the onset of OA. Biomechanical
SIEBELT ET AL
stress induced by strenuous running tended to increase
Hsp90 protein levels in rat articular cartilage. Hsp90
inhibition and the subsequently increased Hsp70 levels
enabled chondrocytes to maintain cartilage homeostasis
by increasing sGAG amounts above baseline in order to
protect the ECM from increasing biomechanical impact
during physical activity. Hsp90 inhibition further improved subchondral bone thickness and reduced synovial macrophage activation. Specific modulation of
chondrocyte Hsp90 activity might prove to be an attractive therapeutic intervention to prevent OA.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors approved
the final version to be published. Dr. Siebelt had full access to all of the
data in the study and takes responsibility for the integrity of the data
and the accuracy of the data analysis.
Study conception and design. Siebelt, Jahr, Groen, Sandker, Waarsing,
Kops, Müller, van Eden, de Jong, Weinans.
Acquisition of data. Siebelt, Jahr, Groen, Sandker, Waarsing, Kops,
Müller, van Eden, de Jong, Weinans.
Analysis and interpretation of data. Siebelt, Jahr, Groen, Sandker,
Waarsing, Kops, Müller, van Eden, de Jong, Weinans.
REFERENCES
1. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and
metabolic abnormalities in articular cartilage from osteo-arthritic
human hips. II. Correlation of morphology with biochemical and
metabolic data. J Bone Joint Surg Am 1971;53:523–37.
2. Odding E, Valkenburg HA, Stam HJ, Hofman A. Determinants of
locomotor disability in people aged 55 years and over: the Rotterdam Study. Eur J Epidemiol 2001;17:1033–41.
3. Sun HB. Mechanical loading, cartilage degradation, and arthritis.
Ann N Y Acad Sci 2010;1211:37–50.
4. Saxon L, Finch C, Bass S. Sports participation, sports injuries and
osteoarthritis: implications for prevention. Sports Med 1999;28:
123–35.
5. Blanco FJ, Guitian R, Vazquez-Martul E, de Toro FJ, Galdo F.
Osteoarthritis chondrocytes die by apoptosis: a possible pathway
for osteoarthritis pathology. Arthritis Rheum 1998;41:284–9.
6. Arya R, Mallik M, Lakhotia SC. Heat shock genes—integrating
cell survival and death. J Biosci 2007;32:595–610.
7. Samali A, Orrenius S. Heat shock proteins: regulators of stress
response and apoptosis. Cell Stress Chaperones 1998;3:228–36.
8. Terauchi R, Takahashi KA, Arai Y, Ikeda T, Ohashi S, Imanishi J,
et al. Hsp70 prevents nitric oxide–induced apoptosis in articular
chondrocytes. Arthritis Rheum 2003;48:1562–8.
9. Galois L, Etienne S, Grossin L, Watrin-Pinzano A, CournilHenrionnet C, Loeuille D, et al. Dose–response relationship for
exercise on severity of experimental osteoarthritis in rats: a pilot
study. Osteoarthritis Cartilage 2004;12:779–86.
10. Etienne S, Gaborit N, Henrionnet C, Pinzano A, Galois L, Netter
P, et al. Local induction of heat shock protein 70 (Hsp70) by
proteasome inhibition confers chondroprotection during surgically
induced osteoarthritis in the rat knee. Biomed Mater Eng 2008;
18:253–60.
11. Grossin L, Etienne S, Gaborit N, Pinzano A, Cournil-Henrionnet
C, Gerard C, et al. Induction of heat shock protein 70 (Hsp70) by
proteasome inhibitor MG 132 protects articular chondrocytes from
cellular death in vitro and in vivo. Biorheology 2004;41:521–34.
12. Takahashi K, Kubo T, Goomer RS, Amiel D, Kobayashi K,
Hsp90 IN OSTEOARTHRITIS
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Imanishi J, et al. Analysis of heat shock proteins and cytokines
expressed during early stages of osteoarthritis in a mouse model.
Osteoarthritis Cartilage 1997;5:321–9.
Siebelt M, Waarsing JH, Kops N, Piscaer TM, Verhaar JA, Oei
EH, et al. Quantifying osteoarthritic cartilage changes accurately
using in vivo microCT arthrography in three etiologically distinct
rat models. J Orthop Res 2011;29:1788–94.
Pap G, Eberhardt R, Sturmer I, Machner A, Schwarzberg H,
Roessner A, et al. Development of osteoarthritis in the knee joints
of Wistar rats after strenuous running exercise in a running wheel
by intracranial self-stimulation. Pathol Res Pract 1998;194:41–7.
Takahashi KA, Tonomura H, Arai Y, Terauchi R, Honjo K,
Hiraoka N, et al. Hyperthermia for the treatment of articular
cartilage with osteoarthritis. Int J Hyperthermia 2009;25:661–7.
Roman-Blas JA, Jimenez SA. NF-␬B as a potential therapeutic
target in osteoarthritis and rheumatoid arthritis. Osteoarthritis
Cartilage 2006;14:839–48.
Rice JW, Veal JM, Fadden RP, Barabasz AF, Partridge JM, Barta
TE, et al. Small molecule inhibitors of Hsp90 potently affect
inflammatory disease pathways and exhibit activity in models of
rheumatoid arthritis. Arthritis Rheum 2008;58:3765–75.
Clarke PA, Hostein I, Banerji U, Di Stefano F, Maloney A,
Walton M, et al. Gene expression profiling of human colon cancer
cells following inhibition of signal transduction by 17-allylamino17-demethoxygeldanamycin, an inhibitor of the hsp90 molecular
chaperone. Oncogene 2000;19:4125–33.
Hamilton KL, Gupta S, Knowlton AA. Estrogen and regulation of
heat shock protein expression in female cardiomyocytes: cross-talk
with NF␬B signaling. J Mol Cell Cardiol 2004;36:577–84.
Roman-Blas JA, Castaneda S, Largo R, Herrero-Beaumont G.
Osteoarthritis associated with estrogen deficiency. Arthritis Res
Ther 2009;11:241.
Tarini VA, Carnevali LC Jr, Arida RM, Cunha CA, Alves ES,
Seeleander MC, et al. Effect of exhaustive ultra-endurance exercise in muscular glycogen and both Alpha1 and Alpha2 AMPK
protein expression in trained rats. J Physiol Biochem 2012. E-pub
ahead of print.
Lundgren K, Zhang H, Brekken J, Huser N, Powell RE, Timple N,
et al. BIIB021, an orally available, fully synthetic small-molecule
inhibitor of the heat shock protein Hsp90. Mol Cancer Ther
2009;8:921–9.
Yun CH, Yoon SY, Nguyen TT, Cho HY, Kim TH, Kim ST, et al.
Geldanamycin inhibits TGF-␤ signaling through induction of
Hsp70. Arch Biochem Biophys 2010;495:8–13.
Glaze ER, Lambert AL, Smith AC, Page JG, Johnson WD,
McCormick DL, et al. Preclinical toxicity of a geldanamycin
analog, 17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (17-DMAG), in rats and dogs: potential clinical relevance.
Cancer Chemother Pharmacol 2005;56:637–47.
Turk MJ, Breur GJ, Widmer WR, Paulos CM, Xu LC, Grote LA,
et al. Folate-targeted imaging of activated macrophages in rats
with adjuvant-induced arthritis. Arthritis Rheum 2002;46:1947–55.
Piscaer TM, Muller C, Mindt TL, Lubberts E, Verhaar JA,
Krenning EP, et al. Imaging of activated macrophages in experimental osteoarthritis using folate-targeted animal singlephoton–emission computed tomography/computed tomography.
Arthritis Rheum 2011;63:1898–907.
Palmer AW, Guldberg RE, Levenston ME. Analysis of cartilage
matrix fixed charge density and three-dimensional morphology via
contrast-enhanced microcomputed tomography. Proc Natl Acad
Sci U S A 2006;103:19255–60.
Aula AS, Jurvelin JS, Toyras J. Simultaneous computed tomography of articular cartilage and subchondral bone. Osteoarthritis
Cartilage 2009;17:1583–8.
Silvast TS, Jurvelin JS, Lammi MJ, Toyras J. pQCT study on
diffusion and equilibrium distribution of iodinated anionic contrast agent in human articular cartilage—associations to matrix
composition and integrity. Osteoarthritis Cartilage 2009;17:26–32.
2111
30. Beckett J, Jin W, Schultz M, Chen A, Tolbert D, Moed BR, et al.
Excessive running induces cartilage degeneration in knee joints
and alters gait of rats. J Orthop Res 2012;30:1604–10.
31. Waarsing JH, Day JS, Weinans H. An improved segmentation
method for in vivo microCT imaging. J Bone Miner Res 2004;19:
1640–50.
32. Botter SM, van Osch GJ, Clockaerts S, Waarsing JH, Weinans H,
van Leeuwen JP. Osteoarthritis induction leads to early and
temporal subchondral plate porosity in the tibial plateau of mice:
an in vivo microfocal computed tomography study. Arthritis
Rheum 2011;63:2690–9.
33. Pastoureau P, Leduc S, Chomel A, De Ceuninck F. Quantitative
assessment of articular cartilage and subchondral bone histology in
the meniscectomized guinea pig model of osteoarthritis. Osteoarthritis Cartilage 2003;11:412–23.
34. Gerwin N, Bendele AM, Glasson S, Carlson CS. The OARSI
Histopathology Initiative—recommendations for histological assessments of osteoarthritis in the rat. Osteoarthritis Cartilage
2010;18 Suppl 3:S24–34.
35. Boehm AK, Seth M, Mayr KG, Fortier LA. Hsp90 mediates
insulin-like growth factor 1 and interleukin-1␤ signaling in an
age-dependent manner in equine articular chondrocytes. Arthritis
Rheum 2007;56:2335–43.
36. Wilson R, Belluoccio D, Bateman JF. Proteomic analysis of
cartilage proteins. Methods 2008;45:22–31.
37. Van der Windt AE, Haak E, Das RH, Kops N, Welting TJ, Caron
MM, et al. Physiological tonicity improves human chondrogenic
marker expression through nuclear factor of activated T-cells 5 in
vitro. Arthritis Res Ther 2010;12:R100.
38. Xing H, Mayhew CN, Cullen KE, Park-Sarge OK, Sarge KD.
HSF1 modulation of Hsp70 mRNA polyadenylation via interaction with symplekin. J Biol Chem 2004;279:10551–5.
39. Hirano M, Shibato J, Rakwal R, Kouyama N, Katayama Y,
Hayashi M, et al. Transcriptomic analysis of rat brain tissue
following gamma knife surgery: early and distinct bilateral effects
in the un-irradiated striatum. Mol Cells 2009;27:263–8.
40. Guo W, Siegel D, Ross D. Stability of the Hsp90 inhibitor 17AAG
hydroquinone and prevention of metal-catalyzed oxidation.
J Pharm Sci 2008;97:5147–57.
41. Soroka J, Buchner J. Mechanistic aspects of the Hsp90 phosphoregulation. Cell Cycle 2012;11:1870–1.
42. Whitesell L, Bagatell R, Falsey R. The stress response: implications for the clinical development of Hsp90 inhibitors. Curr
Cancer Drug Targets 2003;3:349–58.
43. Zhang R, Luo D, Miao R, Bai L, Ge Q, Sessa WC, et al.
Hsp90-Akt phosphorylates ASK1 and inhibits ASK1-mediated
apoptosis. Oncogene 2005;24:3954–63.
44. Pandey P, Saleh A, Nakazawa A, Kumar S, Srinivasula SM, Kumar
V, et al. Negative regulation of cytochrome c-mediated oligomerization of Apaf-1 and activation of procaspase-9 by heat shock
protein 90. EMBO J 2000;19:4310–22.
45. Grad I, Picard D. The glucocorticoid responses are shaped by
molecular chaperones. Mol Cell Endocrinol 2007;275:2–12.
46. Proctor CJ, Lorimer IA. Modelling the role of the Hsp70/Hsp90
system in the maintenance of protein homeostasis. PLoS One
2011;6:e22038.
47. Solit DB, Zheng FF, Drobnjak M, Munster PN, Higgins B, Verbel
D, et al. 17-allylamino-17-demethoxygeldanamycin induces the
degradation of androgen receptor and HER-2/neu and inhibits the
growth of prostate cancer xenografts. Clin Cancer Res 2002;8:
986–93.
48. Shamovsky I, Nudler E. New insights into the mechanism of heat
shock response activation. Cell Mol Life Sci 2008;65:855–61.
49. Zou J, Guo Y, Guettouche T, Smith DF, Voellmy R. Repression
of heat shock transcription factor HSF1 activation by HSP90
(HSP90 complex) that forms a stress-sensitive complex with HSF1.
Cell 1998;94:471–80.
50. Kaarniranta K, Holmberg CI, Lammi MJ, Eriksson JE, Sistonen L,
2112
CLINICAL IMAGES
Helminen HJ. Primary chondrocytes resist hydrostatic pressureinduced stress while primary synovial cells and fibroblasts show
modified Hsp70 response. Osteoarthritis Cartilage 2001;9:7–13.
51. Kimura H, Yukitake H, Tajima Y, Suzuki H, Chikatsu T, Morimoto S, et al. ITZ-1, a client-selective Hsp90 inhibitor, efficiently
induces heat shock factor 1 activation. Chem Biol 2010;17:18–27.
52. Dakappagari N, Neely L, Tangri S, Lundgren K, Hipolito L,
Estrellado A, et al. An investigation into the potential use of serum
Hsp70 as a novel tumour biomarker for Hsp90 inhibitors. Biomarkers 2010;15:31–8.
53. Kimura H, Yukitake H, Suzuki H, Tajima Y, Gomaibashi K,
Morimoto S, et al. The chondroprotective agent ITZ-1 inhibits
interleukin-1␤–induced matrix metalloproteinase–13 production
and suppresses nitric oxide–induced chondrocyte death. J Pharmacol Sci 2009;110:201–11.
54. Neuhold LA, Killar L, Zhao W, Sung ML, Warner L, Kulik J, et al.
Postnatal expression in hyaline cartilage of constitutively active
human collagenase-3 (MMP-13) induces osteoarthritis in mice.
J Clin Invest 2001;107:35–44.
55. Felson DT, Zhang Y. An update on the epidemiology of knee and
hip osteoarthritis with a view to prevention [review]. Arthritis
Rheum 1998;41:1343–55.
DOI 10.1002/art.38019
© 2013 American College of Rheumatology
Clinical Images: Hypercalcemia and miliary sarcoidosis in a 15-year-old boy
The patient, a 15-year-old boy, presented with a 3-month history of nausea, vomiting, fatigue, and worsening headache. Results of
physical examination were notable for enlargement of occipital, cervical, axillary, and inguinal lymph nodes, mild hypertension, and
scattered hypopigmented plaques over the posterior neck and extensor surfaces of multiple joints. Initial laboratory tests revealed
the following values: serum creatinine 2.2 mg/dl, serum calcium 12.9 mg/dl, and ionized calcium 1.74 mmoles/liter. Abdominal
ultrasound demonstrated retroperitoneal lymphadenopathy and medullary nephrocalcinosis. Radiographs of the chest (A) showed
a subtle interstitial prominence, without evidence of hilar adenopathy. Diffuse miliary interstitial nodules throughout the lung
parenchyma were seen on computed tomography of the chest (B). Results of pulmonary function tests were within normal limits.
Additional laboratory studies showed undetectable levels of parathyroid hormone (PTH) (⬍3 pg/ml) and PTH-related peptide (⬍2
pmoles/liter), low 25-hydroxyvitamin D levels (18 ng/ml), and elevated serum lysozyme levels (28 ␮g/ml). Levels of angiotensinconverting enzyme and 1,25-hydroxyvitamin D were within normal limits. The patient underwent transbronchial lung biopsy (C) and
lymph node biopsy (D), which revealed multiple noncaseating epithelioid granulomas (arrows) and prominent areas of hyalinization
(arrowheads). No evidence of mycobacterial or fungal infections was found on serologic tests, histologic stains, or tissue culture.
Sarcoidosis was diagnosed, and initiation of corticosteroid therapy resulted in rapid resolution of hypercalcemia and improvement
of the serum creatinine level within 1 week. Complete resolution of lymphadenopathy was seen after 4 months. To our knowledge,
this is the first description of miliary sarcoidosis in a pediatric patient.
Holly K. Hodges, MD
Pui Y. Lee, MD, PhD
Jonathan S. Hausmann, MD
Lisa A. Teot, MD
Ethan L. Sanford, MD
Kathleen W. Levin, MD
Boston Children’s Hospital
Boston, MA